Power consumption of various architectures

classic Classic list List threaded Threaded
13 messages Options
Reply | Threaded
Open this post in threaded view
|

Power consumption of various architectures

Jan Stary
On Jan 17 20:38:05, [hidden email] wrote:
> > much power with old hardware.
>
> It is not a lot of power; that is a myth.

Having used most of the architectures out there,
can you please say which of them consume most/least power
in regular operation of OpenBSD, if you have such statistics?

(I realize "regular operation" is vague. My only comparison
is the wattmeter on my desk while my machines are on.)

Thank you

Jan

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Miod Vallat
> Having used most of the architectures out there,
> can you please say which of them consume most/least power
> in regular operation of OpenBSD, if you have such statistics?

I can't tell for the exact machines Theo is using, but here are a few
values from my bunker:

- VAX 4000/106 (fast vax, 100MHz processor), quite similar to the one
  Theo is using, two SCSI disks: about 95W.

- Alpha LX164 (2nd generation alpha, 533MHz processor), with a modern (2
  years old) ATX power supply, only one disk: about 85W.

- SGI Fuel (700MHz R16000), original power supply: about 200W.

- 1.4GHz G4 Mac mini: 40W.

- SPARCstation 20, 150MHz RT625 processor, one disk: about 115W.

- Sun Blade 100 (500MHz UltraSPARC-IIe): about 65W.

- HP Visualize B2000 (400MHz PA-RISC): about 130W.

- Lemote Fuloong 2F (800MHz): about 20W.

- Plextor PX-EH160L (landisk): about 15W.

Miod

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Marc Espie-2
On Sun, Jan 19, 2014 at 01:56:05PM +0000, Miod Vallat wrote:
> - VAX 4000/106 (fast vax, 100MHz processor), quite similar to the one
>   Theo is using, two SCSI disks: about 95W.
> - SGI Fuel (700MHz R16000), original power supply: about 200W.
> - HP Visualize B2000 (400MHz PA-RISC): about 130W.

> - Plextor PX-EH160L (landisk): about 15W.

So, imagine, if we ditch VAX, PA-RISC and SGI (425W), we would have enough power
for a cluster of almost 30 landisks !

Nah, just kidding...

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Juan Francisco Cantero Hurtado
In reply to this post by Miod Vallat
On Sun, Jan 19, 2014 at 01:56:05PM +0000, Miod Vallat wrote:

> > Having used most of the architectures out there,
> > can you please say which of them consume most/least power
> > in regular operation of OpenBSD, if you have such statistics?
>
> I can't tell for the exact machines Theo is using, but here are a few
> values from my bunker:
>
> - VAX 4000/106 (fast vax, 100MHz processor), quite similar to the one
>   Theo is using, two SCSI disks: about 95W.
>
> - Alpha LX164 (2nd generation alpha, 533MHz processor), with a modern (2
>   years old) ATX power supply, only one disk: about 85W.
>
> - SGI Fuel (700MHz R16000), original power supply: about 200W.
>
> - 1.4GHz G4 Mac mini: 40W.
>
> - SPARCstation 20, 150MHz RT625 processor, one disk: about 115W.
>
> - Sun Blade 100 (500MHz UltraSPARC-IIe): about 65W.
>
> - HP Visualize B2000 (400MHz PA-RISC): about 130W.
>
> - Lemote Fuloong 2F (800MHz): about 20W.
>
> - Plextor PX-EH160L (landisk): about 15W.

HP PA-RISC J6750 with 2 CPUs, 2 SCSI disks and 2GB of RAM: about 250W

Beagle Bone Black at full speed: 3W

--
Juan Francisco Cantero Hurtado http://juanfra.info

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Артур Истомин
On Sun, Jan 19, 2014 at 09:43:20PM +0100, Juan Francisco Cantero Hurtado wrote:

> On Sun, Jan 19, 2014 at 01:56:05PM +0000, Miod Vallat wrote:
> > > Having used most of the architectures out there,
> > > can you please say which of them consume most/least power
> > > in regular operation of OpenBSD, if you have such statistics?
> >
> > I can't tell for the exact machines Theo is using, but here are a few
> > values from my bunker:
> >
> > - VAX 4000/106 (fast vax, 100MHz processor), quite similar to the one
> >   Theo is using, two SCSI disks: about 95W.
> >
> > - Alpha LX164 (2nd generation alpha, 533MHz processor), with a modern (2
> >   years old) ATX power supply, only one disk: about 85W.
> >
> > - SGI Fuel (700MHz R16000), original power supply: about 200W.
> >
> > - 1.4GHz G4 Mac mini: 40W.
> >
> > - SPARCstation 20, 150MHz RT625 processor, one disk: about 115W.
> >
> > - Sun Blade 100 (500MHz UltraSPARC-IIe): about 65W.
> >
> > - HP Visualize B2000 (400MHz PA-RISC): about 130W.
> >
> > - Lemote Fuloong 2F (800MHz): about 20W.
> >
> > - Plextor PX-EH160L (landisk): about 15W.
>
> HP PA-RISC J6750 with 2 CPUs, 2 SCSI disks and 2GB of RAM: about 250W
>
> Beagle Bone Black at full speed: 3W

How can i find out this numbers? From power block sticker?

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Riccardo Mottola
In reply to this post by Miod Vallat
Hi,

Miod Vallat wrote:

>
> I can't tell for the exact machines Theo is using, but here are a few
> values from my bunker:
>
> - VAX 4000/106 (fast vax, 100MHz processor), quite similar to the one
>    Theo is using, two SCSI disks: about 95W.
>
> - Alpha LX164 (2nd generation alpha, 533MHz processor), with a modern (2
>    years old) ATX power supply, only one disk: about 85W.
>
> - SGI Fuel (700MHz R16000), original power supply: about 200W.
>
> - 1.4GHz G4 Mac mini: 40W.
>
> - SPARCstation 20, 150MHz RT625 processor, one disk: about 115W.
>
> - Sun Blade 100 (500MHz UltraSPARC-IIe): about 65W.
>
> - HP Visualize B2000 (400MHz PA-RISC): about 130W.
>
> - Lemote Fuloong 2F (800MHz): about 20W.
>
> - Plextor PX-EH160L (landisk): about 15W.
>
For some of machines I also have, the numbers are similar. I was running
quite some archs at my home to test stuff!

A Dual-processor SGI Octane was around 175W, lower-end video-card.

I was able to optimize power on several machines by changing/swapping
out video cards. Especially if used only for testing or running
headless, you don't need them Sometimes you save quite a lot of power.
Sometimes you can even remove the video-card.

The second very important step are Hard-disks. Old SCSI hard disks run
hot and consume a lot! Just using newer ones (perhaps with the
Wide-Narrow adapter like in the case of old SPARCStations) help
stressing the system les and often improves even the performance.

I am very tempted to attach IDE hdd's with a SCSI adapter in some old
machines, however I don't know how reliable those SCSI-IDE converters
are.. I might not want to end having problems with them, posts on the
web vary really a lot.

Riccardo

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Sunny Raspet
In reply to this post by Артур Истомин
On Mon, Jan 20, 2014 at 12:30 AM, Артур Истомин <[hidden email]> wrote:
> How can i find out this numbers? From power block sticker?

Absolutely not.  And also not from the manufacturer's specs (for those
architectures and machines that still have this information published
and publicly available.)

The reason for this is, of course, that the actual running load of a
machine is usually considerably less than these numbers.  The x86 Dell
PowerEdge line, for instance, uses about twice the power at startup
that it does in regular operation, which is in turn still not the full
delivery capacity of the hardware; while loading increases this number
somewhat, it's a relatively minor increase that's closer to the
regular operation number than the startup number.

Stickers will only get you the maximum amount the box is *capable* of
drawing.  Which is fine for some things, but not fine for other
things, and not terribly reliable if you're interested in real-world
budgeting per the numbers I mentioned above.  The way to go is to use
a clamp meter, which can be had relatively cheaply.  (But you've
donated to the project first, yes?  Yes.  Perfect.)

-slr

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Juan Francisco Cantero Hurtado
In reply to this post by Артур Истомин
On Mon, Jan 20, 2014 at 05:30:32AM +0000, Артур Истомин wrote:

> On Sun, Jan 19, 2014 at 09:43:20PM +0100, Juan Francisco Cantero Hurtado wrote:
> > On Sun, Jan 19, 2014 at 01:56:05PM +0000, Miod Vallat wrote:
> > > > Having used most of the architectures out there,
> > > > can you please say which of them consume most/least power
> > > > in regular operation of OpenBSD, if you have such statistics?
> > >
> > > I can't tell for the exact machines Theo is using, but here are a few
> > > values from my bunker:
> > >
> > > - VAX 4000/106 (fast vax, 100MHz processor), quite similar to the one
> > >   Theo is using, two SCSI disks: about 95W.
> > >
> > > - Alpha LX164 (2nd generation alpha, 533MHz processor), with a modern (2
> > >   years old) ATX power supply, only one disk: about 85W.
> > >
> > > - SGI Fuel (700MHz R16000), original power supply: about 200W.
> > >
> > > - 1.4GHz G4 Mac mini: 40W.
> > >
> > > - SPARCstation 20, 150MHz RT625 processor, one disk: about 115W.
> > >
> > > - Sun Blade 100 (500MHz UltraSPARC-IIe): about 65W.
> > >
> > > - HP Visualize B2000 (400MHz PA-RISC): about 130W.
> > >
> > > - Lemote Fuloong 2F (800MHz): about 20W.
> > >
> > > - Plextor PX-EH160L (landisk): about 15W.
> >
> > HP PA-RISC J6750 with 2 CPUs, 2 SCSI disks and 2GB of RAM: about 250W
> >
> > Beagle Bone Black at full speed: 3W
>
> How can i find out this numbers? From power block sticker?
>

I use a (good) power meter. Don't buy the cheapest one.

--
Juan Francisco Cantero Hurtado http://juanfra.info

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Christian Weisgerber
Juan Francisco Cantero Hurtado <[hidden email]> wrote:

> > How can i find out this numbers? From power block sticker?
>
> I use a (good) power meter. Don't buy the cheapest one.

From time to time, c't magazine reviews the cheap wattmeters available
on the German (~ European) market.  Their latest test, in the
September 23 issue, revealed a number of models that were "eerily
in agreement with our 3,500-euro reference unit", including the
cheapest one (10 euros).

But really, any unit you buy is probably good enough.

--
Christian "naddy" Weisgerber                          [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Juan Francisco Cantero Hurtado
On Tue, Jan 21, 2014 at 07:31:14PM +0000, Christian Weisgerber wrote:

> Juan Francisco Cantero Hurtado <[hidden email]> wrote:
>
> > > How can i find out this numbers? From power block sticker?
> >
> > I use a (good) power meter. Don't buy the cheapest one.
>
> From time to time, c't magazine reviews the cheap wattmeters available
> on the German (~ European) market.  Their latest test, in the
> September 23 issue, revealed a number of models that were "eerily
> in agreement with our 3,500-euro reference unit", including the
> cheapest one (10 euros).
>
> But really, any unit you buy is probably good enough.
>

The problem isn't the precision of the numbers, the real problem is the
quality of the build and especially the quality of the circuit. I don't
want to set fire to my house :P

--
Juan Francisco Cantero Hurtado http://juanfra.info

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Артур Истомин
On Wed, Jan 22, 2014 at 01:59:51AM +0100, Juan Francisco Cantero Hurtado wrote:

> On Tue, Jan 21, 2014 at 07:31:14PM +0000, Christian Weisgerber wrote:
> > Juan Francisco Cantero Hurtado <[hidden email]> wrote:
> >
> > > > How can i find out this numbers? From power block sticker?
> > >
> > > I use a (good) power meter. Don't buy the cheapest one.
> >
> > From time to time, c't magazine reviews the cheap wattmeters available
> > on the German (~ European) market.  Their latest test, in the
> > September 23 issue, revealed a number of models that were "eerily
> > in agreement with our 3,500-euro reference unit", including the
> > cheapest one (10 euros).
> >
> > But really, any unit you buy is probably good enough.
> >
>
> The problem isn't the precision of the numbers, the real problem is the
> quality of the build and especially the quality of the circuit. I don't
> want to set fire to my house :P

I suspect most wattmeters the world over - cheap noname Chinese devices

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Shawn K. Quinn-2
In reply to this post by Juan Francisco Cantero Hurtado
On Mon, Jan 20, 2014, at 06:22 PM, Juan Francisco Cantero Hurtado wrote:
> I use a (good) power meter. Don't buy the cheapest one.

There's also the possibility of using a clamp-style AC ammeter on the
power cable and multiplying by the nominal line voltage.

--
  Shawn K. Quinn
  [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Power consumption of various architectures

Philip Guenther-2
On Thu, Jan 23, 2014 at 4:49 PM, Shawn K. Quinn <[hidden email]> wrote:
> There's also the possibility of using a clamp-style AC ammeter on the
> power cable and multiplying by the nominal line voltage.

Nope.  http://en.wikipedia.org/wiki/Power_factor


Philip Guenther